The exact age determination of fish is one of the most important elements in the study of their population dynamics. It forms the basis for calculations leading to a knowledge of the growth, mortality, recruitment and other fundamental parameters of their populations.
Many species of fish can be aged from the discontinuities which occur in their skeletal structures. These discontinuities may result from either changes (such as temperature) in the environment which the fish inhabits, or changes (such as spawning) in the physiology of the fish. However, many fish live in such a uniform environment that discontinuities do not form in their skeletal structures and ageing of these fish has to be done indirectly; it may often be impossible. In the first part of this section the methods used for fish with skeletal discontinuities will be described and in the second part the methods available for fish without them. The third part deals with growth rates and the fourth part describes methods of obtaining age compositions from age-length keys.
Almost every skeletal structure has been used for age determination of fish. Of these otoliths and scales are the most widely used because they are easy to collect and store. The thin bones of the head and of the pectoral and pelvic girdles have also been used. Although bones are easy to store dry, they are time-consuming to prepare. The flesh has to be removed by boiling them in water and then the fat removed in a solvent, otherwise they go rancid in storage. Vertebrae are commonly used for rajids, which do not have large bony otoliths and for tuna. Again these need time-consuming preparation. Ray vertebrae have to be cleaned in a boiling solution of sodium hydroxide and then stored in methanol. Tuna vertebrae are stored dry. Spines of some dogfish, Squalus acanthias are also used for regular age determination of this species in England.
When starting an investigation it is worth examining several different structures to see which gives the best results. In some instances two methods will complement each other.
As otoliths and scales are the two structures most widely used for age determination, only methods for reading these will be described. Most of these methods are equally applicable to reading the other structures with either no or slight modification.
The sagittal otolith is that used for age determination of most fish. It is located in the sacculus of the inner ear. The manner of cutting into the head of the fish to remove it depends on the type of fish. It amy also depend upon whether the fish are to be sold subsequently in which case damage to the fish must be kept to a minimum.
For flatfish, a cut in the position illustrated in Fig. 4.1a exposes the otoliths and they are easily withdrawn with forceps (Fig. 4.1c). When otolithing large numbers of flatfish the operation is speeded up by using a board with a slot, 3 x 10 cm, cut in the surface. The point of incision in the head of the fish is laid on the board over the slot; this enables the knife to be driven easily through the skull at point X (Fig. 4.1c), and the cut made cleanly in the correct position (Fig. 4.1b).
A number of methods can be used to extract otoliths from “roundfish” e.g. gadoids, the simplest being a transverse cut across the head, slightly behind the eyes, far enough into the skull to allow the head to be broken open. A knife or scalpel can be used for small fish but a normal metal kacksaw for larger sizes makes the operation quick and easy (Fig. 4.2). An inexperienced worker can easily cut too deeply and damage the otoliths, but the right technique is soon learnt. An alternative method is to lift off the top of the skull. This is a more difficult method, particularly with large fish, although there is less chance of damaging the otoliths (Fig.4.3a). If the otoliths must be removed without apparent damage to the fish then it is possible to extract them from under the gill (Fig.4.3b) or from the roof of the mouth.
The scale lies in a pocket in the skin of the fish and is divided by a horizontal line into two areas. The embedded area is covered with striations and concentric rings, while the exposed area is unstriated. Scales vary in shape depending on the contours of the fish. The best scales for age and growth rate determination are generally to be found on the shoulder of the fish between the head and the dorsal fin (Fig.4.4).
The fish from which scales are to be taken should first be washed under cold running water. During the washing, the body of the fish should be rubbed lightly in a head-to-tail direction in order to remove any loose scales which may have rubbed off other fish. Using forceps, a scale is then taken from the shoulder of the fish. A slight resistance should be felt when it is pulled from its pocket in a head-to-tail direction. If no resistance is felt or if the scale is not observed to pull from its pocket, it is probably that it did not belong to that fish and it should be discarded. When a satisfactory scale has been taken, it should be cleaned by dipping in fresh water and rubbed between thumb and forefinger to remove any dirt or mucus. If sufficiently large, it should then be quickly examined by eye to check that the area covered with concentric rings is not damaged. Sometimes a scale is removed from its pocket at some stage in the fish's life; when this occurs, a new one is grown very quickly, but the new scale has a regenerated centre which makes it quite useless for age determination. Regenerated scales can easily be detected by the confused nature of the striations and the absence of concentric rings near the centre.
Storage systems need to have the following advantages:
(a) Occupy the minimum space.
(b) Need no maintenance, such as the continuous refilling of tubes holding otoliths in a volatile liquid.
(c) Allow easy access and reference.
The most simple way, which can be used for the majority of species, is first to thoroughly clean the otoliths and then to store them dry in conveniently sized paper envelopes or plastic bags which are labelled and stacked in boxes. If it is necessary to keep the otoliths wet, then small sample tubes of the appropriate size should be used and the most suitable liquid selected. Alcohol or glycerine or alcohol/glycerine mixture or creosote may be used, but care should be taken to ensure that the ring structure on the otolith is not damaged or rendered unreadable by storage in an unsuitable medium. Formalin should not be used. Whatever liquid is chosen as the storage medium, it is advisable to check regularly its effect on the appearance of the otolith. In some cases short-term storage may make the rings more easily readable, but this may be followed by a period when there is a progressive deterioration in the clarity of the ring structure. Small otoliths, such as those of sandeels or mackerel, can be conveniently handled and stored by mounting a series in resin on a microscope slide (Fig. 4.5).
The scales should be mounted on glass microscope slides, the same way up as they came off the fish, that is convex side up. A scale is slightly curved in cross section and will curl off the slide if mounted the wrong way up. The scales may stick directly to the slide once they have dried but in difficult cases either a spot of egg albumen or for very thick scales, gelatine containing a fungicide or bactericide (e.g. thymol) placed under the scales will help them to stick. Large scales, e.g. tuna, may be collected dry in small envelopes. When slides are used they should first be numbered in a set order so that it is possible to relate a particular scale with any data that may be recorded. When a slide has its predetermined number of scales on it, the scales should be levelled up one with the other. The line which divides the scale into two parts should be in a horizontal plane in the slide, with the area containing the concentric rings underneath. This will help when the scales are examined under the microscope at a later stage. It is usual to take two scales, one from each side of the fish, one to act as a check on the other (Fig. 4.6).
Methods of viewing otoliths and scales do have some aspects in common but as the otoliths is a three-dimensional structure and the scale almost two-dimensional, the techniques used for otoliths are generally more complex than those required for scales.
As already stated otoliths are three-dimensional structures but they do not necessarily grow at the same rate equally in all dimensions. If there is a pattern in the otolith it will be composed of a number of concentric shells with different radii. Depending on the amount of organic material in each shell or zone, its appearance will vary from extremely opaque to completely hyaline (transparent). When reading otoliths it is usually preferable to identify and count the opaque zones. If any characteristic growth patterns are visible in the otolith they will usually appear opaque zones. In the simplest case one complete cycle of otolith growth (e.g. an annual cycle) consists in the laying down of one opaque zone and of one hyaline zone. The first zone is usually called the ‘nucleus’ of the otolith.
The two types of zones or rings must never be referred to as “light” and “dark”. This only creates confusion, because the method of illuminating the otolith determines whether a zone appears as a dark or light ring (Fig.4.7). The terms “summer zones” and “winter zones” should also be avoided. It is also best to avoid the term annuli unless it is certain that the rings or zones are annual.
The ring structure may be most easily visible immediately the otoliths are removed from the fish. If so, then every effort should be made to read them at that time. For some species it may be the only time at which they are readable.
To view an otolith a number of techniques are available. The simplest of these is to immerse the whole otolith in a clear liquid (water is quite commonly used), illuminate it from above, and view it against a dark background. This method is suitable only if the otoliths are relatively thin and translucent and all the rings can be seen (Fig. 4a). In many species the outer rings become very narrow once the growth rate of the fish slows down. These narrow rings sometimes grow only on the underside of the otolith, and are completely invisible when the whole otolith is viewed from above in the manner described (Fig.4b). They can be seen only when a cross section of the otolith is viewed (Fig. 4c). When investigating any species of fish it is always necessary to check, by examining a cross section, whether these narrow rings are present before accepting an age based on viewing the whole otolith from above. Failure to understand this type of growth pattern in otoliths can result in gross underestimates of age.
It is not possible to see the narrow outer rings more clearly by grinding either surface of the whole otolith and then viewing it from above. Grinding the otolith will make it more translucent but the rings on the underside will still not be visible from above, and grinding may remove some of the rings completely (Fig.4.8d).
When the structure of an otolith makes it necessary to view it in cross-section then a number of possibilities exist, either longitudinal or transverse or diagonal. All these alternatives, illustrated in Fig.4.9, need to be investigated before the most suitable section for routine age determination is chosen. It is frequently found that a very slight change in the angle of a cross-section will help with the interpretation of a difficult otolith.
A simple grinding machine, for making cross-sections only, enables a wide variety of otolith sections to be prepared, with the certainty that each section is across the nucleus (Fig.4.10). After a little practice it is quite simple to break larger otoliths at the required place with the fingers or by holding one end of the otolith firmly in a pair of forceps. If by accident the break occurs off centre, it is a simple matter to correct this by using the grinding wheel. Wetting the surface of the otolith with water or cedar wood oil, applied with a fine paint brush, fills in any irregularities caused either by grinding or breaking and makes the otolith much easier to read.
Whatever method is used to prepare the otolith it is necessary to ensure that the section is taken across the centre of the nucleus. Failure to do this can lead to a number of errors in interpretation. If the break misses the nucleus completely then the first ring will appear to be the nucleus and the age will be underestimated by one year (Fig. 4a). If only the tip of the nucleus appears on the section then the whole appearance and spacing of the rings will be altered, making it easy to misinterpret. Possibly the true nucleus will be ignored, but certainly some confusion will result (Fig.4b). Only when the section passes through the centre of the nucleus will the true size of the nucleus and rings and their relationship to one another be clearly visible (Fig.4c).
To section an extremely small otolith, e.g. that of an eel, the otolith should be placed on a microscope slide and either covered with sellotape or sandwiched in sellotape before being placed on the microscope stage. If illuminated from below it is easy to identify the nucleus and, by pressing on it with a scalpel, to break the otolith directly across it. The sellotape will prevent the very small broken halves from being scattered.
Once a satisfactory section has been prepared it can be viewed either by transmitted light or by illumination from above or after being burned in the manner described by Christensen (1964)
The majority of readers of gadoid fish otoliths in Europe use the transmitted light method of illumination. The prepared section is mounted in plasticine with the surface to be viewed horizontal. The otolith is then illuminated from the side and the sectioned surface placed in shadow. This method offers considerable advantages over the method of direct illumination from above because it allows the reader to see more of the detailed structure of the otolith (Fig. 4.12a).
A simple device used at the Fisheries Laboratory, Lowestoft allows the reader to put the surface of the section into any degree of shadow required, which may well be different for parts of the otolith. This device consists of an adjustable height bar. A brass base plate supports a vertical pillar in which is housed a captive threaded rod. By turning this thread the bar is raised or lowered to the height required (Bedford, 1964) (Fig.4.12b).
Rings on the otolith of some species of fish, e.g. soles (Solea solea), are not visible when viewed by any of the methods already described, and the narrow outer rings and split rings of other species, e.g. plaice (Pleuronectes platessa) and turbot (Psetta marina), often present considerable difficulties. If the prepared cross-section is gently burned over a very low flame of either a small Bunsen burner or a spirit lamp until it is slightly charred, then the appearance of the existing rings is changed and a narrow black ring is produced at each change from hyaline to opaque material reading outwards from the nucleus of the otolith. (Christensen, 1964).
The amount and rate of burning varies between species and the technique required to achieve the best results has to be learnt. Care must be taken to burn the whole of the sectioned surface of the otolith evenly. If the otolith is burned too much it will crumble into a grey ash; this usually occurs first at the edges of the section and it results in the complete disappearance of the ring structure. If insufficient heat is applied then the organic zones will not char.
To achieve the best results the centre of the sectioned surface should be held barely touching the side of the flame and removed when it begins to turn dark brown. If after examination, it is found to require additional burning then it is a simple matter to return it to the flame until the desired result is obtained. After being burned the otolith is picked up on a piece of plasticine which enables it to be moved under the microscope to any desired position.
Each black ring encloses a white area representing the total growth during one year. A large number of extremely fine hair-like concentric black rings can be seen in the white zone, but the true, much thicker black annual ring is clearly distinguished (Fig.4.13).
The use of the burning technique makes it possible to age species whose otoliths were previously impossible to read and to increase the accuracy of age determination of many other species, particularly those of the older fish whose age had been constantly underestimated in the past by other techniques. Problems of nucleus difficulties and false rings can often be resolved after burning. The otoliths of all species do not burn in this way.
Often a good technique is to use a combination of different viewing methods, e.g. whole otolith, illuminated from above against dark background, and sectioned-burned or unburned otolith. Queries arising from one method can sometimes be resolved by using the alternative method on the second otolith of the pair.
Sometimes a particular year will produce unusual growth zones. For example, the opaque zone may either be exceptionally wide or have a check in the middle or have some other unusual characteristic which is easily recognizable.
The examples that have occurred in the otoliths of North Sea plaice were found in the 1944 and 1947 year-classes. Fish from the 1944 year-class often had an extremely thin 1946 opaque zone that was easily recognized; this thin zone sometimes occurred in up to 20 percent of a sample of fish of the 1944 age group (Fig. 4.14). Many of the 1947 year-class fish had a distinctive triple nucleus; this was found in 0 and 1 group fish along the Dutch coast and could be identified many years later in older fish taken from the central North Sea (Fig.4.15).
Although these unusual zones often create difficulties of age determination when they are being formed, once the difficulties have been resolved they serve as a useful check on the validity of the age readings many years later by acting as a biological mark on the otolith. However, one must always guard against the danger of becoming too familiar and thus holding unfound assumptions. Always be objective.
|whole otoliths||direct||water, alcohol xylene||transmitted or direct from above|
|embedding followed by grinding (see ref. below)||dry||direct from above|
|Broken Otoliths||direct||dry, surface is moistened with xylene, cedar wood oil||direct from above or horizontal|
|water, alcohol xylene|
|after grinding||dry, surface is moistened with xylene or with cedar wood oil||direct from above or horizontal|
|water, alcohol xylene|
|burning||dry, surface is moistened with xylene or with cedar wood oil||direct from above|
Otoliths age reading by means of surface structure examination from S.Wiedeman-Smith (1968)
A summary of all the methods of reading otoliths is given in Table 4.1.
Scales are almost two-dimensional structures. The anterior part is formed of a series of sclerites which should extend in a regular pattern from the centre of the scale (Fig. 4.16). If they do not and they are confused or irregular then the scale is almost certainly a replacement scale and should not be used for age determination. The structural discontinuities used for age determination result from irregularities in the pattern of the sclerites; they may be slightly distorted (Fig. 4.16) or they may be slightly closer spaced than the majority of the sclerites; usually the discontinuities are narrow and they are usually called ‘rings’; as stated earlier the term ‘annuli’ should not be used because this presupposes that the rings are annual. Thus the scale presents a different picture to the otolith.
Because scales are thin structures they need no preparation before viewing; the scales should be cleaned before they are stored.
For reading, the slide on which the scales are mounted is placed on the stage of a low-power microscope. The mirror and condenser are removed from the microscope, and light from a lamp is reflected through the scale by means of a piece of white card or reflector. The field of the microscope should appear dark apart from a small segment at the bottom which should be very bright (Fig. 4.17). The magnification used depends upon the size of the scale; in general, the lowest possible magnification is the best because it enables the whole scale pattern to be seen.
Staining can be used to intensify structural discontinuities or to make visible those not available by ordinary light. Galstaff (1952) describes a method of staining tuna vertebrae with alizarin but it takes up to 12 days' preparation. This demonstrates the disadvantage of methods based upon staining; they are too time-consuming for processing large amounts of material, even though batch processing will allow a quicker throughput. However, if staining is the only method which will show discontinuities then it must be used, even if it does result in smaller numbers of age determinations.
Polarized light and phase differentiation are also techniques that can be used.
In this section it is assumed that some pattern of structural discontinuities (for ease of reference termed ‘rings’ for both otoliths and scales, except when referring specifically to one of the structures) exists in the structure which is being used for age determination and that it has been made visible by some technique. The next step is to determine whether any time-scale can be allotted to the pattern of rings. This time-scale need not be annual. There are several ways in which to do this.
(a) by observing the timing of ring formation;
(b) by following a strong year-class through the fishery;
(c) by using the Petersen method.
The opaque zone in the otoliths of North Atlantic species of fish is formed during the period of greatest growth and it is usually wider than the hyaline ring, particuarly during the early years of the fish's life. The hyaline zone is laid down mainly during the period of slowest growth. But the timing of ring formation is considerably more complex than this simple statement suggests.
Although the opaque zone is sometimes referred to as a summer zone, the timing of its formation can vary considerably, depending upon the species of fish, where it lives and its age. In general in the northern hemisphere, the opaque zone starts to grow earliest in the southern part of the species' range and begins later with increasing distance north. Within each area the younger fish begin to lay down the opaque zone before the older fish. In the North Sea the otoliths of a young cod may show the beginning of an opaque zone in February but an older fish may show no signs of it until June. A similar difference in the timing of opaque zone formation between young and old fish also occurs in the north-east Arctic, but the whole process is delayed by about two months; the opaque zone in the older fish may still be visible on the edge in January of the following year. This is illustrated in Fig. 4.18.
The type of zone forming on the edge of the otolith can be recorded from a series of samples taken throughout the year. Then, provided that the zones are annual, the growth pattern shown in Fig. 4.19 should be observed. This series of plaice otolith photographs clearly illustrates both the timing of zone formation and the rate of growth of the otoliths during a year and there can be no doubt that the zones are annual. It should be noted that all these fish belong to the same age group 4, and have been caught during one year in the same area (central North Sea). There can be a wide variation in the timing of the formation of zones from year to year within a single species, between different areas and between fish of different age groups. All these factors must be considered when attempting to collect data to confirm the timing of zone formation.
When the samples are taken from a wide range of age groups from different areas, then the seasonal difference in type of edge is not likely to be as marked. If sufficient otoliths are available, the type of edge present should be recorded separately for each age group from each area.
Exactly the same techniques can be used for determining the time of ring formation on scales. The change in this instance is not between one type of material and another but in the formation of the sclerites.
This method can be validly used whatever the period of the time-scale over which the rings are laid down (for example, six months) as long as the sequence is regular.
If a series of otolith or scale samples are available for a number of years from a stock, it is often possible to confirm the validity of age determinations by identifying a large year-class and following it through the fishery for several years. If the techniques used are sound, then a number of independent readers, each given a different year's otoliths or scales should all produce an age composition showing the dominant year-class; this was the way in which the early North Sea sole otolith readings were confirmed.
This method is named after the Danish fisheries' scientist G. G. Joh. Petersen, who first described it. It depends upon the following sequence of observations:
(a) a length distribution of the species being studied has several modes which are readily separable;
(b) observations of the otoliths or scales of the fish of each length show that almost all the fish which constitute a length mode have the same number of zones or rings on their otoliths or scales.
(c) the reason for the origin of the modes can be determined, enabling a time-scale to be placed on the zones or rings.
The use of this method will be described in more detail in section 4.11.1.
Artificial time markers can be introduced into skeletal structures by injecting chemicals into fish. The initial work was based on the use of lead acetate but this is toxic and tetracycline is now commonly used. It has the advantage of being an antibiotic drug, stable in solid form. It is used in saline solution which must be used immediately or within 24 h if kept in refrigeration.
Tetracycline is readily absorbed by vertebrate animals and deposited in bony structures where calcification is taking place. The areas in which tetracycline is deposited in skeletal tissue fluoresce yellow in ultraviolet light, enabling them to be detected easily.
In teleost fish, which possess acellular bone, the tetracycline is laid down as a narrow ring timing the point of injection to within a month, the time taken to completely excrete excess tetracycline (Fig. 4.20). In elasmobranchs, which have partially calcified, cartilaginous, cellular bone the tetracycline is laid down diffusely throughout the skeleton present at the time of injection. Parts of the skeleton laid down subsequently contain no tetracycline.
Initial tank experiments are carried out, if possible, to determine the optimum dose rate of tetracycline. In the field the fish have to be tagged, as well as injected, because the technique does depend upon retrieving the injected fish so that their otoliths and scales can be examined. In the North Sea the method has been used for cod and whiting (Jones and Bedford, 1968) and for rays (Holden and Vince, 1973).
When it has been possible to view the zones or rings, count them satisfactorily and establish that their formation conforms to a definite time-pattern, then it is possible to age the fish. In the remainder of this section it will be assumed that the fish have one spawning period a year and that the zones on its otoliths (rings on its scales) have an annual pattern of formation.
The terms age, age group and year-class are frequently used. The age of a fish at a given time refers to the period of time from birth to a given point of time. When the age of the fish has been established it can be assigned to the appropriate age group which is an integral number of years, according to a convention based, on an arbitrarily-adopted birthday. Fish said to be of a given year-class are fish born in that particular year.
To simplify aging, age determinations are usually done in terms of age groups with reference to a designated birthday. This birthday does not have to coincide with the biological time of birth. For convenience, there is a convention to use 1 January for most Atlantic species, even though the time of spawning of some members of a species may be in the December of the previous year and other species may not finish spawning until May. This system of identifying fish by age groups is shown schematically in Fig. 4.21 and by photographs of a set of North Sea plaice otoliths in Fig. 4.19. In the latter figure the fish of age group 4 have four opaque and four hyaline zones in January. By June the fifth opaque zone is visible but it does not become a 5-group fish until 1 January 1972. For the whole period 1 January 1971 to 31 December 1971 it is a 4-group fish. Its true age on 1 January 1971 would be 4 years exactly (assuming it was spawned on 1 January 1967) and on 31 December 1971 4 years, 11 months, 31 days.
The great advantage of this method is that it enables the year-class of a fish to be established immediately by subtracting age group from year of capture (in the above example 1971 - 4 = 1967) and it permits easy grouping of data collected by different laboratories. The method does require good knowledge of the pattern of zone or ring-formation because it may be necessary to discount zones or rings that have already formed (the fifth opaque and hyaline zones in the plaice shown in Fig. 4.19 from June to 31 December) or to assume that a zone or ring should have been formed but it is not yet visible. This is shown diagnostically in Fig. 4.18; the opaque-zone of a 15 year old Arctic cod is not visible until the January of the year after which it is laid down; that is, the 1973 opaque zone is not visible until January 1974. In November and December 1973 it is necessary to take this into account and not count the visible hyaline edge as the edge-forming in the 1973–74 winter.
Notations such as age 4 + carry the same meaning if the same systematic convention described above is used but notations such as 4 ++, 4+++ or even 4++++, the last meaning a fish almost 5 years' old, are so meaningless that they have no value. The terminology ‘3 ring fish’, ‘4 ring fish, is also confusing unless it is known that all rings are consistently laid down within a very restricted time period. For example, a species for which 1 January is the official birthday may normally lay down its scale rings in December. A fish chronologically aged 4 years (say, born theoretically on 1 January 1967 and caught on 14 January 1971) will be called a 3 ring fish if it has not laid its fourth ring down by that date.
Some species exhibit a different ring structure after the onset of spawning. In Arctic cod the post-spawning opaque zones are extremely narrow and regular. These post-spawning zones can be identified, counted and the otoliths used to provide data on the number of times each fish has spawned, in addition to giving the age (Rollefsen, 1933) (Fig. 4.22a).
One feature that can be observed on the otoliths of mature cod is the wider and brighter hyaline zones that sometimes occur immediately before the first post-spawning opaque zones.
The otoliths of the same species of fish from different areas are often quite different in appearance, reflecting the different growth pattern of the fish. An experienced otolith reader is often able to identify, with considerable precision, the origin of a sample of otoliths taken from an area where a mixing of stocks occurs. Three distinctive cod types from the north-east Arctic north of Norway are illustrated in Fig. 4.22.
(a) from the eastern Barents Sea;
(b) from the western part of the area; and
(c) the resident coastal type.
Types (a) and (b) are found in members of the stock which have lived in different parts of the north-east Arctic; type (c) is from a different stock the Norwegian coastal stock.
Differences in the size of growth zones on scales (distance between two rings), particularly the first growth zone, are used to separate stocks of North Sea herring. Also the country of origin of North Atlantic salmon can be determined by the number of growth zones laid down while they are in freshwater during the early part of their life when the growth rate is much slower under these conditions than in the sea (Fig. 4.23).
There can be differences in the number of rings present in the two otoliths taken from the same fish. Fortunately this is an extremely rare occurrence and can usually be detected by a fairly superficial examination of both otoliths, when a deformity in the shape or structure of one or both otoliths is easily visible. Scales rarely differ unless one is a regenerated scale or unless they are taken from different parts of the body.
It is not at all uncommon for the growth rates of the same species of fish taken from the same area to show an extremely wide variation. For example, the ages of small North Sea plaice in the 5 cm size group 25–29 cm can range from 1 to 7, the 1 group being most likely to occur in December when their actual age is 1 year and 11 months' old. The 5 cm group 30–34 cm could contain fish between 2 and 12 years' old, and among the larger fish (45–49 cm in length) the age could lie between 7 and more than 20 years.
Wide variations in growth rates also occur in other species and great care should be taken not to fall into the error of first looking at the length of the fish from which the otoliths have been taken, deciding what the age is likely to be, and then looking at the otolith or scale and attempting to make the preconceived age fit the visible structure, either by grouping or omitting rings.
A similar error is to attempt to define size limits within which the nucleus or zones must fall, based on the size of zones found on other otoliths. A slow-growing fish may have three or four annual zones occupying the same space as one or two zones on a fast-growing fish (Figs. 4.2.4 and 4.2.5). It is therefore impossible to say that some zones must be false because three or four narrow annual zones occupy the space of one very wide annual zone. These comments apply also to scales.
EITHER THERE ARE RECOGNIZABLE RINGS THAT CONFORM TO A REGULAR TIME-SCALE ON THE OTOLITH OR THERE ARE NOT. LENGTH IS NOT A CRITERION OF AGE.
The main source of difficulty in otolith reading is to distinguish as certainly as possible the true annual zones from secondary, false, split or check zones. If the otolith burns satisfactorily this usually clears up any queries. It is usually assumed that any ring on a scale which runs round the whole of the anterior edge is a ‘true ring’ and is counted and that one which does not do this is a ‘false ring’ and is not counted. Solutions to these problems can be gained only by study of the species under investigation.
Age determination is a skill which has to be learnt. Some people become more proficient at it than others; some never become competent. While the European Inland Fisheries Advisory Commission (EIFAC) is preparing a scale atlas based on fish of known age, the whole purpose of reading scales and otoliths is to determine the age of fish whose age is unknown and whose absolute age will always be uncertain. The best reader in the world cannot state with 100 percent certainty the exact age of a wild fish.
To avoid errors and systematic bias, frequent cross-checking of results between readers and by the same reader repeating the reading of a sample at a later date will ensure that no wide differences of interpretation occur and that the readings remain consistent. Readers must always be objective.
Any major change in the environment in which a fish lives is likely to produce ring formation (in this sub-section ring and zone formation will be considered synonymous). In the earth's temperate zones differences between summer and winter are marked by both changes in water temperature and amounts of food available. It is these very regular and very marked changes which make the age determinations of most temperate fish species so easy. But equally marked changes occur within the tropics. Many rivers in the tropics are subject to seasonal floods. During the floods food is abundant and the fish grow rapidly. In the dry season food becomes very scarce and the fish often starve. This results in very marked rings forming on the scales. Examples are the Gambia (Svensson, 1933) and the middle Niger (Daget, 1956). In Lake Chad (Hopson, 1965) has described how the fall in temperature in November-January causes the formation of rings on the scales of Lates niloticus. Poinsard and Troadec (1966) describe the age determination of Pseudotolithus senegalensis and P. typus, two West African marine species, both of which lay down two opaque zones during two warm seasons and two hyaline zones in two cold seasons. Mature fish of the former species spawn twice a year.
Garrod (1959) aged Tilapia esculenta in Lake Victoria by proving that the rings which occur at the edge of the scale were laid down at spawning. A time-scale was allotted to the rings by then proving that this species spawned twice a year in the north of Lake Victoria and once a year in the south. (Seshappa, 1969) describes a similar study to that of Garrod, using the rings caused by spawning that form on the scales of Indian mackerel Rastrelliger kanagurta. Seshappa also examined the otoliths of this species but found no zones on them.
When faced with the age determination of a new species the first question must be ‘Has it got rings on its scales or zones on its otoliths which can be read by some method?’. The second question must be ‘What causes these rings or zones so that a time-scale can be set to them?’. The regularity of the event causing ring formation must be checked to determine that no other events will lay down rings similar to the main seasonal rings. This is most likely to happen if the species being studied is living in a small volume of water, such as a small lake or pond. In temperate zones the main rings may be caused by summer-winter variations in the environment but other events may produce rings identical to annual ones. These events might be either plankton blooms or sudden unseasonal changes in temperature or accidental discharge of pollutants, to quote three possible examples.
However, some fish live in uniform environments, notably in the polar and tropical regions and consequently no rings whatsoever are laid down in their skeletal structures.
This method has already been referred to in section 18.104.22.168. In that section it was referred to as a means of age validation of fish which had discontinuities in their skeletal structures. The method can be used only with species which have a restricted spawning season so that the fish bred in a single season can be identified as a single mode in a polymodal length distribution. The mode with the lowest value is identified as soon as possible after spawning; these will be O-group fish. Subsequent modes will be 1-group, 2-group fish and so on. The method can be very good for young fish but becomes increasingly less useful for older fish as the growth rate slows down and the modes merge.
In practice length-frequency distributions of fish are plotted caught over the shortest time period possible; the shorter the time period the more precisely the modes will be defined. A regular sequence of such length frequency distributions enables the progression of the modes to be followed.
The following examples illustrate the problem.
(a) Sardinella aurita
In this tropical species age determinations by otoliths and scales is not easy but the length frequency distribution ages those with a mode at 11.5 cm as O-group and those with a mode at 18.5 cm as 1-group.
Fig. 4.26. Sardinella aurita. Ghana, Feb. 1968. Length distribution in the purse-seine landings.
(b) The Greenland halibut (see next sheet)
Tagging does not enable individual fish to be aged unless the age of the fish at tagging is known indirectly, for example, by using hatchery reared fish or known 0-group from the Petersen method. Otherwise, all that is known is that recaptured tagged fish must be at least as old as the number of years for which they have been at liberty. Growth parameters can be calculated from the recaptures of tagged fish (Gulland and Holt, 1959). The method described in this paper is very useful for fish living in areas where the growth is continuous throughout the year, because it is based on this assumption. It is less useful where this assumption does not hold, but it can be used if large numbers of fish recaptured at annual intervals are available (Jones and Jonsson, 1971).
However, changes in the growth rate caused by both the tag and the tagging procedure often occur. Therefore, the growth of the recaptured fish cannot be taken as a representative of that of unmarked fish in the population in the same period.
(b) The Greenland halibut (Reinhardtius hippoglossoides Walb.)
|The age of this Arctic species can not be determined by the use of otoliths and scales, but the length distribution gives a good estimate of the length of O-, I-, II- and III-group fish. Furthermore, it is possible to follow the growth of the year-classes during the year (data from Smidt, 1969).|
|Fig. 4.27.||The Greenland halibit (Reinhardtius hippoglossoides Walb.) Godthåb, 1953–63. changes in the length distributions during the year.|
As the scale grows at approximately the same proportional rate as the fish, it can be used for growth rate determination. This relationship, which may be linear or curvilinear, can be determined by plotting scale size (Vn defined below) against fish length. If it is linear the length of the fish when each ring was laid down on the scale (referred to as 11,12,13…1n) can be calculated using the formula
where 1n is the length of the fish when the n.th ring is laid down
Vn is the length of the striated part of the scale to the n.th ring
V is the total length of the striated portion of the scale
L is the total length of the fish
c is a constant
If c = 0 or is small, values of 1, 12, etc., can be calculated easily as follows.
To calculate this relationship automatically a simple proportioning board can be used (Fig. 4). It consists of a graduated rule fixed to the right-hand side of a polished board at right angles to the bottom edge. This edge is lipped. An ungraduated rule, pivoted at the bottom left-hand corner of the board and long enough to overlap the top right-hand corner, completes the device.
To use the board an image of a scale is projected onto a rectangular piece of white card. The bottom end of one edge of the card is placed on the nucleus of the projected scale, and the positions of the rings and the edge of the scale are marked on the edge of the card at an angle of 90° to the baseline. The marked card is then positioned on the board with the edge of the card corresponding to the nucleus flush with the lip at the bottom of the board. The lower edge of the diagonal, ungraduated scale is then placed at that length on the graduated rule at the right-hand edge of the board which corresponds to the length of the fish. The card is then moved along so that the line which marks the edge of the scale is directly under the lower edge of the ungraduated rule. This rule is then pivoted and placed directly over each of the marks on the card and the appropriate length is read each time from the graduated rule on the right.
Griffiths (1968) describes a more sophisticated electronic proportion device which automatically presents the growth data on the scale of a digital voltmeter.
As c becomes increasingly greater or smaller than zero so the values of 11, 12, etc., become increasingly incorrect using this method, the error being greater for 11 and decreasing progressively toward 1n.
It is much less easy to back-calculate lengths from otoliths than from scales. In whole otoliths the edges of the zones are rarely well-defined and in sectioned otoliths the same difficulty exists. In addition there is the difficulty of sectioning the otolith accurately across the middle of its nucleus. As already described in section 22.214.171.124. the relative size of the zones depends upon the point at which the section is made. Small differences in the point at which the section is made may be perfectly acceptable for age determination, but can result in a considerable difference in the size of the zones that are visible for measurement. A series of transverse sections taken through the nucleus of the same otolith will show considerable differences in both the shape of the otolith and the size of the rings (Fig. 4).
If measurements are made on otoliths they are taken with a calibrated graticule fitted in the eyepiece of the microscope. Projection techniques cannot be used because the intnsity of illuminations is too low.
The proportion of different age-groups pf fish in a catch or in the population is called its age composition. If the problem is to get an estimate of the age composition of a single catch of one species this is simply done by determining the age of all the fish in the catch. Table 4.2 shows the age composition of 200 fish of one species taken in a single trawl haul; the age of every fish caught was read.
The results in the table can be shown in a histogram (Fig. 40.30).
Fig. 4.30. Age composition of a hypothetical catch.
In the previous example the age composition of a single, small catch was determined but normally that of the total landings is required. Every year this may total several million fish of each species landed and obviously it would be impossible to age each fish individually. Instead, an age-length key is constructed and used with the length composition data to estimate the total age composition.
The principle behind constructing an age-length key is that the length range occurring in the landings is split into length strata of equal length range, for example, 5–9 cm, 10–14 cm, 15–19 cm or 3–4 cm, 5–6 cm etc. The length range of each stratum will depend upon the growth rate of the species. For slow-growing fish the strata need to be smaller than for fast-growing fish. Usually a maximum of twenty length strata is sufficient.
Having decided upon the length strata a certain number of otoliths (scales, bones) are collected from randomly-taken fish in each stratum. How the number of fish taken is calculated is described in section 126.96.36.199. The ages of these fish are determined and an age-length key constructed. An example is shown in Table 4.3, which is for whiting caught off Scotland.
The length strata (or length groups as they are described in the table) are 2 cm. A total of 377 whiting were aged of which 15 fish were in the 28–9 cm length stratum, 22 fish in the 30–1 cm stratum and so on. In the total example 46 fish were aged 2 years' old, 60 fish 3 years' old etc. However, these fish were selected by strata so the age data in the bottom row do not represent the age composition of the catch. They must be combined with the total numbers of fish landed (Table 4.4).
From Table 4.3 we know how the age groups are distributed in the different length groups. In the length group 28–9 cm 13 fish out of 15 (=86.67%) were two years' old and 2 fish (= 13.33%) were three years' old. (From now on it is assumed that the percentage of each age group in each length stratum is a valid sample applicable to the landings). Therefore, 86.67% of the total number landed in the 28–9 cm group (8,100 fish) are two years' old, and 13.33% are three years' old, i.e. 86.67.8100/100 = 7,020 fish are two years' old and 13.33 .8100/100 = 1,080 fish are three years' old. The calculated values for all other length groups are seen in Table 4.4. The sum of each column gives an estimate of the total number landed in each age group.
Calculation is easier if a raising factor is used for each length group. This is obtained by dividing the total number landed by the total number aged for each length stratum. For the 28–9 cm group this is: .8100/15 = 540. Multiplying this factor by the number of fish of each age in the age length key gives the number of fish in each age group, e.g.: 540.2 = 1,080. The final result is shown in Table 4.4.
If mean lengths for age are required for constructing growth curves the age composition data (Table 4.4) MUST be used, NOT the age-length key (Table 4.3).
The results in Table 4.4 are showas a histogram in fig.4.31
Fig.4.31. Age composition of Whiting. Hebride, January-March 1968,trawl.
Such histograms can be obtained for a number of years and it is possible to put all these figures together. Fig.4.32 shows the age distribution in the landings of sole for a number of years on the west coast of Jutland. The black column in the figure correspond to large “years-classes”. (In some species there are very grate fluctuations in the annual recruitment and there is a corresponding variation in the proportions of the age-group in the population. In the Danish sole fishery the large year-classes are those of1958 and 1963 and they can be followed in the catches for several years. Such large year-classes are the basis of many fisheries and their absence can cause a serious reduction in the catch.
Fig.4.32. Sole, North Sea, are composition of the landing in Hvide Sande, 1961–70
The number of fish sampled for age is a question of money and time. The precision will be improved by any increase in sampling but the time (and money) spent will increase too. A balance between the precision required in these estimates and the time used to get this precision has to be struck.
The variance in each length group of the estimated numbers of fish age i can be written down:
The coefficient of variation of Ni is:
where in a certain length stratum
N = number of fish in the landings
n = number of fish sampled for age
pi = number of fish in the age-length key>br< at age i, in fractions
n.pi = number of fish in sample, age i
Ni = estimated number of fish, age i, in the landings of any one size group
Details of the derivation of these formulae are given by Gulland (1966).
The variance decreases by 1/n as more fish (= n) are aged. But it increases as the number of age groups in each length stratum (pi) increases. If there was only a single age group in a length stratum pi = 1, (1-pi) = 0 and the variance would be zero.
It is usual to try to make the variance in each length stratum equal so the answer to the question of ‘how many fish in an age length key?’ is as many as possible with the manpower available, the total number of otoliths being divided between strata on the basis of the number of age groups likely to be present in each stratum. This has to be determined for each species. In brief it means ageing more large fish, where age for length usually varies considerably, than smaller fish, where age and size are closely related.
Different kinds of gear are often used in one fishery. For example, in the Danish sole fishery two kinds of gear are used, trawls with small meshes and set nets with larger meshes. Therefore, the landings from the set net fishery contain more big fish than the trawl landings. These differences are visible in Fig. 4.3. As it is complicated to obtain a sample which gives a representative picture of the length distribution of the total catch it is easier to distinguish between and keep the two sets of data apart. One age-length key can then be used separately with each set of length composition data and two sets of age compositions calculated. Also, if the fisheries with different gears take place on separate parts of the same stock, for example, one fishery on young fish and another on the older fish, it is preferable to compile separate age-length keys for each fishery.
Growth rates of male and female fish of the same species may differ; for example, if the females are growing faster than the males a three-year-old female could be bigger than a male of the same age. Handling the two sexes together in the age-length key can give unwanted complications. It is, therefore, necessary to separate the two sexes in constructing age-length keys. It is not always possible to sex commercially-landed fish so this method is not always usable because the method also requires length compositions by sex.
As fish grow during a year how often should age-length keys be compiled if they are to be valid? For slow-growing fish (less than 10 cm a year) annual age-length keys are used but for faster-growing fish it is necessary, to use keys compiled over shorter periods. Table 4.5 shows an example for North Sea cod.
|I. Quarter||III. Quarter|
|II. Quarter||IV. Quarter|
The smaller (and younger) sizes in particular show large changes in abundance during the year. In the first quarter the number of fish in the 30–39 cm length group is 37 and in the fourth quarter the number has increased to 64. During the year fish smaller than 30 cm have grown and have successively entered the 30 cm length group. most of the change in the distribution is due to the entrance of one-year-old fish in the last two quarters.
A single age-length key would give a biased age composition especially if there was a difference in the number of fish landed in each quarter. In this instance it is better to use the quarterly age-length keys with the respective quarterly landings.
Very often several countries carry out research on the same species and to facilitate cooperation and to ensure that only the most important and useful results of such investigations which are published standard tables are indispensable.
An example has already been given in section 3 (Appendix 5) for length composition data; this also includes age composition data.
Bedford, B. C., 1964 Two mechanical aids for otolith reading. Res.Bull.ICNAF, (1):79–81
Christensen, J. M., 1964 Burning of otoliths, a technique for age determination of soles and other fish. J.Cons.Perm.Int.Explor.Mer, 29(1):73–81
Daget, J., 1956 Mémoire sur la biologie des poissons du Niger moyen. 2. Recherches sûr Tilapia zillii (Gerv). Bull.Inst.Fr.Afr.Noire (a), 18(1):165–223
Galstaff, P. S., 1952 Staining of growth rings in the vertebrae of tuna, (Thunnus thynnus). Copeia, 2(1952): 102–5
Garrod, D. J., 1959 The growth of Tilapia esculenta in Lake Victoria. Hydrobiologia, 12(4):268–98
Griffiths, P. G., 1968 An electronic fish scale proportioning system. J.Cons.Perm.Int.Explor. Mer, 32(2):280–2
Gulland, J. A., 1966 Manual of sampling and statistical methods for fisheries biology. Part 1. Sampling methods. FAO Man.Fish.Sci., (3):pag. var.
Gulland, J. A. and S. J. Holt, 1959 Estimation of growth parameters for data at unequal time intervals. J.Cons.Perm.Int.Explor.Mer, 25:47–9
Holden, M. J. and M. R. Vince, 1973 Age validation studies on the centra of Raja clavata using tetracycline. J.Cons.Perm.Int.Explor.Mer, 35:13–7
Hopson, A. J., 1965 Winter scale rings in Lates niloticus (Pisces, Centropomidae) from Lake Chad. Nature, Lond., 208(5014):1013–4
Jones, B. W. and B. C. Bedford, 1968 Tetracycline labelling as an aid to interpretation of otolith structures in age determination - a progress report. ICES CM 1968/GEN: 11:3 p. (mimeo)
Jones, B. W. and J. Jónsson, 1971 Coalfish tagging experiments at Iceland. Rit.Fiskideildar., 5(1), 27 p.
ICES, 1970 Demersal species. Nominal catch and fishing effort stock record data, 1968. Stat. News Lett.ICES, (46):154 p.
Poinsard, F. and J-P. Troadeo, 1966 Détermination de l'âge par la lecture des otoliths chez deux espèces de Sciaenidae cuest africaines (Pseudotolithus senegalensis C. et V et Pseudotolithus typus Blks). J. Cons. Perm. Int. Explor. Mer, 30 : 291–307
Rollefsen, G., 1933 The otoliths of the cod. Preliminary report. Fiskeridir. Skr. (Havunders), 4(3) : 14 p.
Seshappa, G., 1969 The problem of age determination in the Indian mackerel, Rastrelliger Kanagurta, by means of scales and otoliths. Indian J.Fish., 16(1–2):14–28
Svensson, G.S.O., 1933 Freshwater fishes from the Gambia river (British West Africa). Results of the Swedish expedition 1931. K.Sven.Vetenskapsakad.Handl., 12(3):1–102
Smidt, E., 1969 The Greenland halibut Reinhardtius hippoglossoides (Walb), biology and exploitation in Greenland waters. Medd.Dan.Fisk.Havunders., 6(1–4):79–148
Wiedeman-Smith, S., 1968 Otolith age readings by means of surface structure examination. J.Cons.Perm.Int.Explor.Mer, 32(2):270
Fig. 4.1 Removal of otoliths from flatfish.
Fig.4.2 Removal of otoliths from roundfish.
|Fig. 4.3||Alternative methods of removing otoliths from roundfish: (a) by lifting off the top the of the skull; (b) by extracting them from under the gills.|
Fig. 4.4. Hatched areas show the best positions for taking scales
Fig. 4.5 Storage of small otoliths by mounting them on a microscope slide.
Fig. 4.6 Method of mounting scales on a glass slide
Fig. 4.7 An otolith viewed by two different methods.
Fig. 4.8 Plaice otoliths: (a) a thin otolith from a young fish, (b)a thick otolith from an old fish.
Fig. 4.9 A series of cross sections taken through a cod otolith.
Fig. 4.10 A grinding machine for preparing otolith sections.
Fig. 4.11 Three transverse sections taken at various points through the same otolith.
Fig. 4.12 Method of mounting and illuminating the cross section of a cod otolith.
Fig. 4.15 Plaice otoliths, all with a similar nucleus.
Fig. 4.17 The lighting of the microscope field for scale examination.
|Fig. 4.18||Zone formation of the otoliths of North sea cod (upper graph) and Arctic cod (lower graph)|
|opaque - - - hyaline|
|Fig. 4.20||Otoliths of a west of Scotland cod, released after tagging and injection of tetracycline, June 1968; recaptured July 1969.|
Fig. 4.21 The age of fish. taking 1 January as the birthday.
Fig. 4.22 Typical north – east Arctic cod otolith types.
Fig. 4.23 Growth curve for salmon.
Fig. 4.24 Plaice otoliths from similar sized fish.
Fig. 4.25 Two cod otoliths from fish of the same size but widely differing age.
Fig. 4.28 A proportioning board.
Fig. 4.29 A series of transverse sections taken through the nucleus of the same otolith